Evolving design :).

[DavidS]

I am glad I went with a self designed 3D Printer.My design continues to evolve as I continue to use it.I began with a very simple design that used a floating Y rail, with two rails (the entire rail system for the Y axes moving up and down along the Z axis). Now I have the X axis rails suspended above the bed, a single Y rail, and the print bed is lowered away from the extruder head instead of the extruder head lifting. It is much more stable with the new configuration, and is easier to expand in the future for larger print volumes.Then there is the controller. I began by using the GPIO pins on a Raspberry Pi Model B, hooked up to three ULN2803's, with a few voltage regulators and a 12 volt power supply. I found that to be a problem so I moved to a Arduino, that had its own set of issues. So I dug up one of my old Parallax Propeller P8X32A boards, and I am now using the Propeller for the controller, still using ULN2803's for the current sinks, and the extra voltage regulators, along with a 12V power supply for the motors and heating element.So designing a 3D printer from scratch has taught me enough to know how to do it correctly. I now have a very stable, reliable 3D printer (still adjusting for scale [only 4 hours since the last revision]).And if I had began with the design I ended up using I would have saved $40, making the total cost around $140. If it were a commercial product with a production cost of $150, and about half an hour of labor to construct copies (repetition reduces time to do well), that would be $140 + $10 for labor, + 20% profit equals $180 for a fair market price. And we all know that buying in bulk reduces the price of parts, so if some one did mass produce this design (at least 10 units per run), it would cost closer to $150 retail, for a fair market price.I would like it if all those that do produce 3D printers to sale to remember what it actually costs. And if you are thinking about getting into the market, I would recommend that you do like I have so far and design your own, write all the firmware, test, modify, test and modify some more. Then produce your first run, keeping the profit labeled at $10 labor per unit, plus the cost of parts materials, and utilities used for the application, plus a 20% market up.I will never get into selling them, as I do not have the greed to want to make more money, nor do I have the resources to support a customer base.

[LambdaFF]

It's admirable to be able to build as you do.

As a manufacturing proffessionnal though, I would say that 10$ for labor is low. Unrealistically so. I have experienced first hand that paying low $ makes things more expensive : you can expect more defects which lead to increased inspections and a standing team to analyse and correct.

Successful businesses have a good tech support and a comprehensive warranty. This doesn't come cheap either.

And if you want to remain in business, you need to devote some time to improvements : you'd need to factor that in.

[DavidS]

It's admirable to be able to build as you do.As a manufacturing proffessionnal though, I would say that 10$ for labor is low. Unrealistically so. I have experienced first hand that paying low $ makes things more expensive : you can expect more defects which lead to increased inspections and a standing team to analyse and correct.Successful businesses have a good tech support and a comprehensive warranty. This doesn't come cheap either.And if you want to remain in business, you need to devote some time to improvements : you'd need to factor that in.

Having been down the business road before:Improvements come out of the 20% profit margin.The $10 per hour labor is factored for a one person business, as it would be for the first year or so while raising capital by selling the product. Then if you wish to expand do so carefully, do not make the mistake of MakerBot.Tech support is the primary reason I would never sell such a product. I am no good with social situations.Any higher mark up or labor price is an indicator of either greed, or to ambition of a strategy. Do not kill your company before it is off the ground.

[soofle616]

I'm curious if your numbers include ALL components used or just the ones you didn't already have laying around. The propeller board lists for 34 bucks, cheapest steppers i've seen are around 15 bucks so 45 for 3 of them, and you're already up to 80 dollars before you ever get to the rails, belts, drive gears, hot end, frame, bed, wiring, power supply, etc. Not bashing you AT ALL here so please don't take this personally, it just bothers me when i see claims of a 3d printer or scanner or whatever that someone built for 20 dollars or something otherwise unheard of before. In every one of those stories is the same underlying theme: "i had tons of these expensive purchased parts lying around already so i bought some cheap fixings and built a printer for next to nothing." The assumption is that everyone has the same pile of parts laying around which is simply not true for most people. Good on you and the other makers for the work you've done, i'm just curious how realistic it is that someone else could replicate the build for the same budget.

[curious aardvark]

so what does it look like and what does it print like ?

as I do not have the greed to want to make more money,

what's greed got to do with it ?
I'd just like lots more money :-)

[DavidS]

I'm curious if your numbers include ALL components used or just the ones you didn't already have laying around. The propeller board lists for 34 bucks, cheapest steppers i've seen are around 15 bucks so 45 for 3 of them, and you're already up to 80 dollars before you ever get to the rails, belts, drive gears, hot end, frame, bed, wiring, power supply, etc. Not bashing you AT ALL here so please don't take this personally, it just bothers me when i see claims of a 3d printer or scanner or whatever that someone built for 20 dollars or something otherwise unheard of before. In every one of those stories is the same underlying theme: "i had tons of these expensive purchased parts lying around already so i bought some cheap fixings and built a printer for next to nothing." The assumption is that everyone has the same pile of parts laying around which is simply not true for most people. Good on you and the other makers for the work you've done, i'm just curious how realistic it is that someone else could replicate the build for the same budget.

I do understand where you are coming from, no offence taken.

Steppers $12 for one Nema 17, and $5 each for three small steppers (the small ones are all that is needed for X/Y and feeding the filament). Steppers from Adafruit.

Propeller Proto Board USB, $20 each (if you purchase them 5 at a time, as I always have).

ULN2803's, $2 each, and I am using 3.

Linear bearings, $5 each (could print them for less than a dollar each if I had access to another 3D printer during construction [and use BB's for the balls]).

8mm steel rod (purhased in 6 foot length) for rails, $8.

Wood for frame, print bed etc, less than $7.

Steel and NiCrome to make hot end, $5.

Screws less than $1.

Epoxy less than $1.

Voltage regulators, less than $1 (bulk from Digikey).

Some of the wood was used for the driving spools, and fishing line is used instead of drive belts.

And a few nuts and some all thread for the z drive, $10.

That is less than $120. Though beings that I am half asleep at the moment I may have left some things out, so the $140 I mention above is likely closer to the truth.

Note that I did make some changes to the design as I went along, so I spent over $200 on it as some things got replaced, do to design changes.

So yes I included everything.

As to aesthetics, not that great, though could be made to look better with out any additional cost. I always count everything, including the purchase cost of those things I already had laying around.

[DavidS]

so what does it look like and what does it print like ?



what's greed got to do with it ?
I'd just like lots more money :-)

It looks OK, and it prints great. It does as well as any of the under $500 I have seen used by friends.

The only issue so far is with the firmware, that is the scale is of (I made a misscalc in writing the firmware), that is easy enough to correct (doing so by trial and error takes some time though). And there are a few G-Code commands that it does not know that do get sent from a slicer.


Once the firmware is completely debugged it is good, as that is a one time job.

[DavidS]

My point here is that a good design can be had for little cost, if you take the time to think things through. If I had known the problems with the first design before hand I would have spent less than $140 total (including the price of the things I already had), and had a great 3D printer.

I did not know the problems that I would have along the way. Using a Raspberry Pi was less than ideal, I learned this, using an AVR is nearly as bad as using a Raspberry Pi, I learned this. The original layout of the rails, and steppers was less than ideal, I learned this.

So even if you are learning as you go it still does not cost much. And I hope others can learn from my experience, and save themselves some money, it only costs $140 to build your own if you do it correctly the first time.

I do not understand why people use AVR's for 3D printers, yes it can do the job, though it is way less than ideal. The 32 GPIO combined with 8 cores at 20MIPS each of the Parallax propeller works so much better. And $20 for a propeller Proto Board USB is great for a MCU board that has everything needed (except for the ULN2803's and the stepper voltage regulators).

[soofle616]

I do understand where you are coming from, no offence taken.

Steppers $12 for one Nema 17, and $5 each for three small steppers (the small ones are all that is needed for X/Y and feeding the filament). Steppers from Adafruit.

Propeller Proto Board USB, $20 each (if you purchase them 5 at a time, as I always have).

ULN2803's, $2 each, and I am using 3.

Linear bearings, $5 each (could print them for less than a dollar each if I had access to another 3D printer during construction [and use BB's for the balls]).

8mm steel rod (purhased in 6 foot length) for rails, $8.

Wood for frame, print bed etc, less than $7.

Steel and NiCrome to make hot end, $5.

Screws less than $1.

Epoxy less than $1.

Voltage regulators, less than $1 (bulk from Digikey).

Some of the wood was used for the driving spools, and fishing line is used instead of drive belts.

And a few nuts and some all thread for the z drive, $10.

That is less than $120. Though beings that I am half asleep at the moment I may have left some things out, so the $140 I mention above is likely closer to the truth.

Note that I did make some changes to the design as I went along, so I spent over $200 on it as some things got replaced, do to design changes.

So yes I included everything.

As to aesthetics, not that great, though could be made to look better with out any additional cost. I always count everything, including the purchase cost of those things I already had laying around.

Fair enough, thanks for the cost breakdown and good job building something so inexpensively.